This chapter aims to summarise the different earthworm populations observed between seasons in both a pasture field, located near the river Loddon in Reading, England, and an arable field, located near the river Swale in Holme-On-Swale, England. Each of these fields had sites within them that were subject to differing levels of flooding. In the pasture field, there were three regions: regularly flooded, fast-draining; regularly flooded, slow-draining; and rarely flooded. The arable field was more complex, and could be divided into: rarely flooded crop and margin; occasionally flooded crop and margin; regularly flooded crop and margin; and field and river sides of the river bank.
As the methods listed generated large amounts of data, this chapter has been split into multiple component parts.
Chapter 3: This section consists of the abstract, scope of the chapter, introduction, methods, and
hypotheses that apply to both the pasture and the arable site
Chapter 3A: This section consists of the results, discussion, and conclusion for the pasture field.
Chapter 3B: This section consists of the results, discussion, and conclusion for the arable field.
Chapter 3C: This section discusses the similarities and differences observed between the pasture
and the arable field, and considers the key factors driving the earthworm population in both.
3.3. Introduction
The UK climate is changing. Flooding events associated with increased rainfall have been increasing in both occurrence and intensity (Prudhomme et al., 2003) and the mean annual floodwater discharge in the UK increasing by approximately 12% between 1960 and 2010. While these events can cause catastrophic damage to urban regions they are also affecting arable and pasture regions of the UK, leading not only to losses of crops and livestock, but also to long term
2014 affected over 13,000 ha of agricultural land, reducing winter cereal crop yields by 20%, leaving 50% of winter crops unviable, and reducing grassland availability for grazing by an estimated 30% (ADAS, 2014). With the threat of flooding increasing on agricultural land, one question that arises is what will be the impact of these flooding events on soil fauna – particularly earthworms?
Earthworms are an important organism in the soil ecosystem. They are a key food source for many animal species: nocturnal foragers such as badgers (Skinner and Skinner, 1988) and foxes (Macdonald, 1980), birds (Ausden et al., 2001; Wilson et al., 1999), and soil dwelling fauna such as moles (Funmilayo, 1979) consume earthworms as a greater or lesser part of their diet. Earthworms have also been referred to as ‘ecosystem engineers’ (Jones et al., 1994): organisms which, through their behaviours, cause changes to abiotic or biotic materials, altering the availability of resources to other organisms (Lawton, 1994). Earthworms fulfil this role in the soil environment by their behaviours; for example, their tunnelling increases soil porosity and air spaces (Stork and Eggleton, 1992), consumption of the soil and organic matter contribute to the nutrient turnover of the soil, and casting of digested material increases the aggregate stability of the soil (Zhang and Schrader, 1992). Given that earthworms play such an important role in the soil, it is important to consider whether changes in flooding regimes with changing climatic conditions will impact earthworm populations, and the further implications this may have on crop yields or grassland production.
It has long been an accepted observation that earthworms emerge from the soil after heavy rainfall. However, the precise reason for earthworm emergence remains debated. The early theory that the earthworms were already in poor condition and emerged in order to die (Darwin, 1881) seems unlikely, but other theories have been suggested. Perhaps the most commonly accepted theory is that the earthworms are ‘escaping drowning’. As all earthworm species use the skin as the organ of oxygen transfer, a more accurate description here may be that the earthworms are escaping suffocation. When soil floods, the available oxygen in the soil is rapidly depleted, falling to anoxic levels within a few hours (Ponnamperuma, 1984). With no available oxygen, the earthworms may flee flooded soil in order to avoid suffocation. As a short-term survival strategy, fleeing an anoxic environment may be very successful for earthworms, provided they can move rapidly enough to entirely escape a flooded region. However, over multiple repeated flooding events, this tactic may result in reduced earthworm populations, with earthworms on the soil surface vulnerable to
inundation may cause physical and chemical changes within the soil that create an environment that is either unsuitable for earthworms, such as the reduced oxygen conditions, or which favours one particular ecotype or behavioural subtype over another, such as changes in soil pH, as discussed in Section 2.6. Population fluctuations in regularly flooded regions, therefore, may depend on a number of factors: how likely earthworms are to survive on the surface and repopulate the flooded regions; how viable earthworm cocoons and juveniles remain during and after a flooding event, and whether earthworm species belonging to different ecotypes respond in the same way.
Within arable soils, the role of earthworms becomes even more important; the changes that earthworms cause in soil through their behaviour as ecosystem engineers are vital for crop growth (Bertrand et al., 2015). They alter the physical environment of the soil, increasing soil water infiltration rates (Ernst et al., 2009) and increasing soil aggregate stability (Maeder et al., 2002). Earthworms also alter the nutritional content of soil. They free nutrients essential for crop growth such as nitrogen from soil, either through excretion of consumed soil matter (Barley and Jennings, 1959) or through release from tissue on earthworm death (Syers and Springett, 1984), while their casts contain high quantities of macro- and micronutrients necessary for plant growth and metabolism (Tomati and Galli, 1995). Through their burrowing, casting, and excretion of nutrients, earthworms create a soil environment more conducive for plant growth (Scheu et al., 1999; Tomati et al., 1988), in turn increasing crop yield by 25% when soil nitrogen is otherwise limited (van Groenigen et al., 2014). However, in arable soils earthworm populations are also greatly reduced in comparison to pasture soils (Curry et al., 2002; Boag et al., 1997) due to a number of factors including crush or cutting damage from agricultural machinery (Boström, 1995; Tomlin and Miller, 1988), the use of pesticides (Pelosi et al., 2013; Ball et al., 1986) and the generally low organic matter contents found in arable soils being unsuitable for large earthworm populations (Reeleder et al., 2006).
This chapter aims to address the differences in earthworm populations in sites of a pasture field and an arable field that are subject to different flooding regimes. The chapter will examine the differences first independently, in Chapters 3A and 3B, and then broadly across the whole dataset, in Chapter 3C. The chapter will address these aims through measures of earthworm abundance, biomass, and diversity, alongside measures of environmental variables that may also change with different flooding regimes.
Five broad hypotheses were considered for this chapter:
1. Soil environmental factors will vary with both the time and the flooding regime. Factors such as soil temperature and soil moisture content will vary seasonably, while factors such as soil carbon and nitrogen content will likely be affected by the flooding regime. As discussed in Section 2.5.2, regular flooding may cause soil pH to be less acidic, may lead to reductions in soil bulk density, and may lead to the accumulation of organic matter in soils.
2. Populations of earthworms found in different regions of the field would be greater in the autumn than in the summer, and would be lower in the regularly-flooded regions than in the rarely flooded regions of the field (Section 2.7; Plum 2005).
3. Earthworm populations in regularly-flooded regions are dominated by endogeic earthworm species, as these are generally the most resistant to flooded conditions (Section 2.6; Roots, 1956; Zorn et al, 2008).
4. The abundance and biomass of individual adult earthworms of different species differs according to the flooding regime, with the abundance of species of adult anecic and epigeic earthworms and the biomass of epigeic adults lower in the regularly-flooded areas than in the rarely flooded (Section 2.6; Klok et al, 2006) and differs over the sampling timepoints, with fewer endogeic earthworms in the warmer months due to aestivation.
5. Earthworm population variables are correlated with soil environmental factors. Earthworm populations may be positively associated with factors such as soil carbon and nitrogen percentage, and negatively correlated with soil moisture.
3.4. Methods
3.4.1. Field sites
3.4.1.1. Pasture
The pasture field is located at grid reference SU 75153 68746 near Reading, England. The field runs adjacent to the river Loddon, which regularly escapes its banks during the winter period (T Sizmur, Department of Geography and Environmental Science, University of Reading, 2016, personal communication, 4th October). The site also regularly experiences groundwater flooding, with both fluvial and groundwater flooding events often occurring concurrently and sometimes
field were chosen that were known to undergo different flooding regimes due to the topography of the field (Fig. 3-1). Site 1 regularly floods, but drains rapidly. Site 2 regularly floods, but does not drain rapidly, and is often under standing water for some time following flooding events. Site 3 very rarely floods.
The pasture site was visited every three months over a period of eighteen months, from November 2016 to February 2018. On each visit, six samples were randomly taken in each flooding regime
Figure 3-1. A. Google earth image (screen captured 08/01/2020) of the three sampling sites within the pasture field bordering the river Loddon. Site 1 regularly floods but
drains rapidly. Site 2 also regularly floods but drains slowly, and is often under standing water. Site 3 very rarely floods. B. Ordnance survey map of the same field at
the 1:25 00 scale, with the approximate locations of the sampling sites marked with red pins.
A
3.4.1.2. Arable
The arable field is located at grid reference SE 36200 81600 near Holme-On-Swale in Yorkshire, England. The field runs adjacent to the river Swale, and is regularly inundated during the winter (M Marley 2017, Holme House Farm, Holme-On-Swale, personal communication). Eight regions of the field were chosen that are subject, not only to different flooding regimes due to the topography of the field, but also to different treatments of crops or field margin (Fig. 3-2).
Sites 1, 3, and 5 are regions of the cropped soil that never flood, occasionally flood, and regularly flooded respectively. Sites 2, 4, and 6 are subject to the same flooding regimes, but within the field margin. Sites 7 and 8 are the field and river side of the riverbank that is separated from the field by a fence, and not farmed or accessed by machinery. The bank lies between the field and the river Swale. A phone interview with a representative of the Environment Agency suggested that the bank has likely been there for a period greater than 100 years, and was a dumping location for silt from river dredging to keep the river channel clear (O Saunders, Environment Agency, personal communication, 01/08/2018).
Figure 3-2. A. Google earth image (screen captured 08/01/2020) of each of the 8 sites in the arable field bordering the river Swale Sites 1 and 2 rarely flood. Sites 3 and 4 occasionally flood. Sites 5 and 6 regularly flood. Sites 7 and 8 are the field and river sides
of the bank, respectively. B. Ordnance survey map of the same field at the 1:25 00 scale, with the approximate locations of the sampling sites marked with red pins.
The arable field was visited approximately every three months, from April 2017 to January 2018. Due to poor weather conditions, some of the sampling during the January 2018 period was delayed. The decision to only sample for the year once sampling had been completed in each season, a spring, summer, autumn and winter sampling, was due to the generally low earthworm abundances across all sites. On each visit, six samples were randomly taken in each flooding regime region, giving a total of forty-eight samples per visit.
A
3.4.2. Earthworm sampling
Samples were taken by digging a pit measuring approximately 20 cmx 20 cm x 20 cm. The soil was extracted using a sharp levering motion with a spade in order to prevent earthworm escape, and the soil mass was rapidly removed from the soil into a high sided tray. The extracted soil was hand sorted for live earthworms, while any earthworms living deeper within the soil were extracted using one litre of 0.13 ml L-1 concentration allyl isothiocyanate in deionised water (Zaborski, 2003; Pelosi et al., 2009), which was poured into the pit and left for 30 minutes to drain into the soil. Emerging earthworms were rinsed with deionised water, and stored separately from earthworms collected from the pit. Earthworms were collected live and transported back to the laboratory in damp soil. The soil temperature at 5 cm and 10 cm depths for each pit was recorded by inserting a soil temperature probe into the intact soil adjacent to the pit. A soil sample was collected by hammering a bulk density ring of volume 63.62 cm3 (height 4 cm, diameter 5.5 cm) into the side of the freshly dug pit, approximately 10 cm below the soil surface, taking the measurement between approximately 8 cm and 12 cm soil depth, before the addition of the allyl isothiocyanate solution. The sample was brought back to the lab for analysis of soil moisture content, bulk density, soil pH, and soil carbon and nitrogen content.
In the laboratory, live earthworms were identified using the OPAL “Key to Common British Earthworms” (Jones and Lowe, 2016) and weighed. As only adult earthworms can be identified using that key, juvenile earthworms were weighed and categorised by their colour and size. Juvenile earthworm categories were: greater than 1 gram, as any juveniles greater than a gram in weight were likely to belong to one of two species of earthworm; Lumbricus terrestris and
Aporrectodea longa, as both L. terrestris and A. longa achieve mass > 1 g within 12 weeks of
hatching (Lowe and Butt, 2002); unpigmented and less than 1g; pigmented and less than 1 g; and green and less than 1 g. The majority of juveniles were beneath 1 g in weight. . Green juvenile earthworms were likely to only belong to endogeic earthworms, as they were likely juveniles of
Allolobophora chlorotica. Earthworms below 1 g in biomass and unpigmented in colour were
considered likely to be endogeics. Earthworms below 1 g in biomass and pigmented, such as a dark brown or red colour, were considered likely to be epigeics. Earthworm fragments or dead earthworms were recorded as such, along with their weight and any colour that could be determined.
3.4.3. Soil analysis
Soil samples were dried at 105°C for 24 hours. The weights of the soils were recorded before and after drying. The difference in the weights was used to calculate the gravimetric moisture content of the soil samples, and the weight of the dried soil sample was used in conjunction with the known volume of the sample ring used in the field in order to calculate the bulk density in grams per centimetre cubed (g cm-3) of oven-dried soil. There were no large stones found in the bulk density sampling at either site. The dried soil samples were stored for future analysis.
Soil pH was determined by adding 40 ml of deionised water to 10 g of the dried soil sample in 50ml Fisher centrifuge tubes. These tubes were shaken for two hours and left to stand for one hour in order to allow any particulate matter to settle. Soil pH readings were taken using a Thermo Orion 420A plus pH/ISE Meter, calibrated with pH 4, pH 7 and pH 10 buffers.
Soil carbon and nitrogen content was determined by grinding a subsample of the dried soil in a ball mill. Approximately 100 mg (c. 95 to 105 mg) of the dried, finely ground soil sample was then analysed in a Vario Macro C/N analyser, which gave the percentages of carbon and nitrogen in the sample and the C:N ratio. The C/N analyser was calibrated using samples of glutamic acid in the same weight range as the dried soil. A certified organic analytical standard of Peaty soil from Elemental Microanalysis Ltd (B2176 – batch 133519), gave recoveries of 97% (± 2.21) and 100% (± 2.94) percent of 15.95% C and 1.29% N respectively.
Multiplying the soil carbon percentage by 1.72 gives an estimate of the soil organic matter content, based on the principal that 58% of the soil organic matter is composed of carbon. However, as this conversion factor is not suited to all soil types (Pribyl, 2010), the obtained soil carbon values were not converted into soil organic matter and instead the soil percent carbon and nitrogen values obtained in this analysis were used.
3.4.4. Data analysis
Data were analysed using R version 3.4.2.The environmental factors used in further analysis were: the mean of the soil temperature at 5 cm and 10 cm (°C), soil pH, soil moisture content (%), soil bulk density (g cm-3), soil carbon percentage, and soil nitrogen percentage. The separate datasets of soil temperature at 5 cm and 10 cm were combined into one mean value.
extraction was calculated for each pit. Partial earthworms were not included in this calculation. The abundance was then divided by 0.04 to convert the values from the 20 cm x 20cm x 20 cm area to an area of 1 m x 1 m to a depth of 20 cm. Secondly, the total biomass of earthworms(g m- 2) was calculated by summing the biomass of each individual and partial earthworm body fragments and treating the values in the same manner as the total earthworm abundance; dividing by 0.04. Thirdly, the Shannon diversity index (Equation 3-1) was calculated for each pit. When calculating the diversity index, juveniles were classified as separate species based on the categories listed above. Partial or dead individuals, while included in the total biomass data, were not included in the diversity data. The Shannon index was chosen as the diversity index for this study as it is widely accepted to be more robust with smaller populations potentially dominated by only a few species than indices such as the Simpson Diversity Index (Morris et al., 2014), an advantage when some pits contained very low numbers (i.e. one individual).
𝐻 = ∑(𝑃𝑖) × ln (𝑃𝑖)
Equation 3-1. The formula for determining the Shannon Diversity Index value for each species, where H = Shannon Diversity Index, and Pi = the proportion of the total sample represented by species i.
Datasets were tested for normality and heteroscedascity, and non-normal datasets were transformed. For the both the pasture and arable sites, no transformations were performed on the on the environmental datasets. Within the pasture dataset, the total biomass of earthworms was square root transformed, while the Shannon diversity index values were raised to the power of two. Within the arable dataset, the total biomass of earthworms was square root transformed, while the Shannon diversity index values were log10 +1 transformed. For the other data analyses performed, statistical tests were used which did not require normality, or were non-parametric equivalents to parametric tests.
Repeated measures ANOVAs were not used in the statistical testing of the hypotheses as the same site was being measured, but not the same pit. As the pits were at least 1 m apart, they were considered sufficiently independent to not require a repeated measures ANOVA.